In order to provide the end user with the best equipment available, the staff at the Facility perform a range of tests and calibrations on the systems and their accessories throughout the year. It is also possible to request a specific calibration for spectral radiance or irradiance when the systems are to be used in a unique configuration. This level of service helps maintain the performance of the equipment and ensures that only valid calibration data are supplied to the users.
The Facility has a darkroom equipped with a variety of transfer standards. These include a spectral radiance sphere source, and a spectral irradiance lamp standard with current regulated power supplies. The radiance standard is calibrated by the UK National Physical Laboratory (NPL) from 300 to 2500nm, certificate reference number 2013030091-RB4-13-1A with the irradiance standards calibrated from 250 to 2500nm, certificate number 2014040396-IB4-14. A precision photometer is also used to independently monitor the stability of these standards.
Reference reflectance standards are used extensively by the remote sensing community and the Facility maintains over fifteen Spectralon reflectance targets ranging in size from 5 to 30 cm square. Throughout their life the panels become contaminated or damaged and the Facility has recently developed an in-house process to refurbish and clean the panel surface. Thereafter, the panels are measured and calibrated with reference to the NPL calibrated Spectralon panel transfer standard. We also have a range of brand new, NPL calibrated, spectralon panels dedicated to calibration and QA which never leave the dark room.
The handheld Microtops sun photometers are returned to the manufacturer for annual calibration. The Cimel sun photometer is now part of the Aeronet group and is annually recalibrated by NASA Goddard Space Flight Center at Greenbelt, Maryland, USA.
Every field spectroradiometer is calibrated with each of its field limiting fore optics. The FSF field spectroradiometers are supplied with a choice of fibre optic light guides and fore optic field-of-view lens attachments ranging from 1° to 18°. Each configuration requires its own unique calibration file.
The calibration procedure involves viewing the Hoffman Engineering LS-65-8D source at the NPL calibration level. The luminance level is also verified with an independently calibrated Macam L203 precision photometer. After the spectroradiometer foreoptic is aligned to the output port of the spectral radiance transfer standard, a series of 30 spectral raw data files are recorded. At this stage, tests are also carried out to check for system repeatability. The spectral average of these files is normalized to unit time and gain before being ratioed with the NPL spectral radiance calibration data file. This calibration file is supplied to the user with the FSF post processing Excel Template for post processing spectral radiance measurements.
A specially profiled acrylic or PTFE diffuser accessory is fitted to the input of fibre optic light guide when measuring spectral irradiance. This configuration ensures the system's angular response approximates to Lambert's cosine law. The limited transmission of the acrylic material reduces the spectral range of the spectroradiometer in the ultra-violet and shortwave infrared regions.
The front surface of the diffuser assembly is carefully aligned and positioned at 50cm and 100cm from the NPL calibrated 1KWatt FEL lamp transfer standard. An Optronic Laboratories OL 83A precision current regulated power supply is used to set the FEL lamp current to 8.000A ± 0.02%. As with the radiance calibration the Macam L203 photometer independently verifies that the illuminance level is correct (within limits of uncertainty). A series of measurements are taken to verify system repeatability and finally 30 spectral raw data files are recorded. The spectral average of these files is normalized to unit time and gain before being ratioed with the NPL spectral irradiance calibration data file. The calibration file is supplied to the user with the FSF post processing Excel Template for irradiance measurements.
The calibration of the spectroradiometer's wavelength scale is often over looked by the users. However it is relatively easy to verify and does not always require expensive equipment. The Field Spectroscopy Facility uses a mercury argon spectral line source with emission lines in the ultra-violet, visible and near infrared spectral regions. We also use a calibrated Mylar sheet with absorption lines between 1.1 and 2.5µm, an erbium doped Spectralon panel with absorption features between 350 and 1500nm and a McCrone glass filter with absorption lines across the visible, near infrared and shortwave infrared spectrum. The standard specification limit for a diode array spectrometer would be ± 0.5 to ± 1nm times the wavelength (pixel) resolution. The GER 1500 (GER 3700 Vis/NIR) pixel resolution is nominally 1.5nm with the GER 3700 SWIR1 & 2 pixel resolutions at 6.5nm and 9.5nm respectively. With the ASD FieldSpec Pro FR systems, the data is always interpolated to 1nm, and the wavelength accuracy specification limits are set to ± 1.5nm. When a spectroradiometer fails to meet its wavelength specification limits, new polynomial coefficients or wavelength scales can be calculated to replace the existing calibration data.
The Labsphere Spectralon reference reflectance panels are typically supplied with an 8° - Total hemispherical calibration from the manufacturer. This calibration configuration does not identify any anisotropy in the panel reflectance and may underestimate the reflectance when used under field conditions. An alternative approach is to calibrate the panels with a 0 / 45° geometry and to account for any deviation from a Lambertian reflectance (Jackson, 1992).
Calibration of the reference panels requires a stabilised tungsten-halogen source to be placed at 0° (normal incidence to the panel surface) and a spectroradiometer with narrow field of view fore optics at 45° to the normal. Using the comparison technique the panel under test is spectrally compared to the NPL calibrated transfer standard panel. The ratio of these measurements is then used to spectrally scale the NPL reflectance data to create a new calibration file for the panel under test. A goniometer is used to monitor and compare the Lambertian properties of the Spectralon panels, whether they are new, used or refurbished. This data is compared to a calibrated panel measured in the field goniometer system detailed by Walter-Shea, et al (1993) and the Implementation of a Sun Angle Correction Factor (See Post processing tips, Rollin, 1999) and the FSF Excel templates for reflectance data files.
Jackson R.D., Clarke T. R. and Moran M.S. 1992. Bidirectional Calibration Results from 11 Spectralon and 16 BaSO4 reference reflectance panels, Remote Sensing of the Environment, Vol 40, p. 231-239
Rollin E.M., Emery D.R., Milton E.J. 1997. Reference Panel Anisotropy in Field Spectroscopy, Seventh International Synposium on Physical Measurements and Signatures in Remote Sensing, Rotterdam, Guyot,G. and Phulpin,T. ed.
Walter-Shea E.A., Hays C.J. and Mesarch M.A. 1993. An Improved Goniometer System for Calibrating Field Reference Reflectance Panels, Remote Sensing of the Environment Vol 43, p. 131-138